What this tool does differently
Most compressors give you "low, medium, high" and hand back whatever comes out. That is fine until a government portal, an exam application or a job form says maximum 200 KB and rejects everything above it.
This one works backwards from the size you need. But the more useful difference is what happens when your target is impossible: it tells you before it compresses, rather than handing you an unreadable file and letting you find out after you have submitted it.
Why most compressed scans come out unreadable
Nearly every browser-based PDF compressor works the same way: it renders each page to a picture and saves that picture as a JPEG.
JPEG was designed for photographs. It assumes smooth gradients — skin, sky, grass. Black text on white paper is the opposite: hard edges everywhere. Squeezed hard, JPEG smears a grey halo around every letter, and at small sizes the letters dissolve into each other.
For a scanned document this tool does what an office scanner or a fax machine does instead. It converts the page to sharp black-and-white at a higher resolution, rather than blurry grey at a lower one. Text is black marks on white paper, so storing 256 shades of grey wastes the space you could have spent on detail.
On a real four-page scanned report, the difference measured like this:
- The usual approach needed 111 KB and the text was not readable.
- This approach produced 70 KB and the text was readable.
Smaller and clearer, from the same document, at the same time.
What the tool tells you before you commit
When you add a file, it is examined on your device — no upload, no rendering, just a quick read of its structure. From that it knows whether you have a scan or a normal PDF, and how many pages.
That is enough to mark each size option honestly:
- Recommended — the smallest size that stays comfortably readable
- May soften — it will work, but check it before sending
- Hard to read — this will probably not be legible, and here is what to do instead
For a scanned document, roughly 35 KB per page keeps text comfortable. Below about 21 KB per page it becomes a struggle whatever tool you use. Those numbers come from measuring real documents, not from a formula.
When compressing is the wrong answer
If your form demands 50 KB and you have a five-page colour scan, no compressor on the internet will give you a readable result. That is not a limitation of this tool — it is roughly 10 KB per page, which is not enough pixels to form letters.
The fix is upstream: rescan in black-and-white at 200 DPI. A scanner discards the paper colour, the shadow from the book spine and the sensor noise, all of which are consuming your file size without carrying any information. The same document rescanned properly will compress to your target and still be readable.
No competitor will tell you this, because it ends with you leaving the page. It is still the right answer.
Normal PDFs have a floor
If your PDF was made by Word, Google Docs or a report generator, its text is stored as instructions rather than pixels. Those stay perfectly sharp however much the file shrinks — but they also have a real size that cannot be reduced further.
When a file reaches that floor, this tool stops and says so. It could force the file smaller by converting the text into a picture, but that would destroy exactly the quality you were trying to keep. You will see the smallest sharp version instead, clearly labelled — not a failure message.
Practical notes
- Keep your original. Always. Compression is one-way.
- Open the result before you submit it. Ten seconds now beats a rejected application.
- Leave a margin. If a portal says 200 KB, target 180 KB. Some validators measure slightly differently, and a file that is exactly at the limit sometimes fails.
- Large files are fine. There is no size cap, because there is no upload to cap. The only limit is your device's memory.